CN110574272B

CN110574272B - Matrix converter control method and system

Info

Publication number: CN110574272B
Application number: CN201780082564.6A
Authority: CN
Inventors: 曼朱沙·维贾雅戈帕尔; 塞萨尔·阿曼多·席尔瓦希门尼斯; 李·恩普林厄姆; 利利亚娜·维多利亚·德利洛
Original assignee: ITT Manufacturing Enterprises LLC
Current assignee: ITT Manufacturing Enterprises LLC
Priority date: 2016-12-05
Filing date: 2017-10-24
Publication date: 2023-04-14
Anticipated expiration: 2037-10-24
Also published as: RU2019120515A3; US20190348921A1; EP3549251A1; GB2557294B; GB201620647D0; US11509231B2; CA3046129A1; RU2019120515A; US20210273573A1; EP3549251A4; GB2557294A; JP7066711B2; JP2020501494A; WO2018104808A1; CN110574272A; KR102462047B1; US11923780B2; MX2019006502A; BR112019011578A2; CN115603592A

Abstract

The invention provides a method for generating a control strategy based on at least three switch states of a matrix converter. The at least three switch states are selected based on at least the predicted output current and the desired output current associated with each switch state. In particular, a mathematical transformation of the output current and the output current associated with each of the plurality of switch states is desired for identifying the appropriate switch state.

Description

Matrix converter control method and system

Technical Field

The present invention relates to the field of matrix converters, and more particularly to the field of control methods for matrix converters.

Background

A matrix converter is typically a single-stage AC-AC converter that uses a switch array to convert a first AC signal (of any number of phases) to a second AC signal (of any number of phases) having any amplitude and frequency. One advantage of the matrix converter is that it does not require any large number of energy storage elements.

A typical matrix converter requires that each switch in the switch array be a bidirectional switch capable of blocking voltage and conducting current in both directions. A two-diode two-transistor bidirectional switch is a known method of independently controlling the direction of current flow within a matrix converter.

One known modulation technique for matrix converters uses Space Vector Modulation (SVM) to perform the modulation of the first AC signal. Several SVM techniques are known to those skilled in the art, such as the tri-zero, two-zero, and one-zero methods.

Examples of matrix converters and SVM techniques can be understood with reference to EP 1311057 A1.

Disclosure of Invention

The invention is defined by the claims.

According to an embodiment, there is provided a method of generating a control strategy for a multiphase output matrix converter, the matrix converter being operable in a plurality of switching states, the method comprising: obtaining a target output transform result representing a mathematical transform result of a desired multiphase output current of the matrix converter; identifying a plurality of switch states of the matrix converter; for each of the identified plurality of switching states, obtaining a predicted output transformation result representing a mathematical transformation result of the output current predicted for the switching state; identifying at least three switch states from the plurality of switch states, wherein the location of the target output transformation result is encompassed by an area defined by locations of predicted output transformation results associated with the at least three switch states when mapped using a Cartesian coordinate system; and generating a control strategy for the matrix converter based on the at least three switching states.

Embodiments thus provide a method of generating a control strategy for a matrix converter based on a predicted output current of the matrix converter. The possible switch states of the matrix converter are each associated with a respective switch state, wherein each switch state may be associated with a respective predicted multiphase output current. Mathematical transformations (e.g., alpha-beta transformations) of both the multiphase output currents and these predicted output currents are desired for identifying which predicted output currents, and thus which switch states, are used in the control strategy.

Specifically, at least three of the switch states are identified, wherein a mathematical transformation of the three switch states defines a region containing the mathematical transformation of the desired output current.

Thus, the method according to embodiments selects which switch states to use in the control strategy based on at least the mathematical transformation of the predicted output currents associated with each switch state and the mathematical transformation of the desired output currents of the matrix converter.

Embodiments allow enhanced control of the output current of a matrix converter with high fidelity and reliability. Using the predicted output current to determine the switch states used in the control strategy enables a fast and accurate determination of the control strategy to be maintained.

For each of the identified plurality of switching states, obtaining a predicted output transformation result may include: for each switching state, a predicted output transform result is obtained from a simulated or mathematical model of the matrix converter and a load of the matrix converter.

Accordingly, embodiments may include querying or otherwise determining a predicted output current associated with one or more switch states from a simulated or mathematical model (and associated loads) of the matrix converter. In an embodiment, the simulation or mathematical model may be a table or data set that includes a predicted output current (or more preferably, a mathematical transformation of the predicted output current) for each switching state of various different possible loads and/or input currents. Other simulation or mathematical models, such as circuit simulation software packages, will be apparent to the skilled person.

For each of the identified plurality of switching states, obtaining a predicted output transformation result optionally comprises: predicting an output current of the matrix converter associated with each of the identified plurality of switching states using a simulated or mathematical model of the matrix converter and a load of the matrix converter; and performing a mathematical transformation on the predicted output current associated with each switching state to obtain a predicted output transformation result for each of the identified plurality of switching states.

Identifying the at least three switch states may include: for each of the identified plurality of switch states, obtaining a predicted output transform result error representing a prediction error between a predicted output transform result associated with the switch state and the target output transform result; and identifying at least three switch states, wherein, when mapped using a cartesian coordinate system, the location of the origin is encompassed by an area defined by the locations of predicted output transform result errors associated with the at least three switch states.

Therefore, a method of generating a control strategy based on a prediction error between an output current predicted by a matrix converter and a desired/target output current is proposed. In particular, a prediction error between a predicted output current and a desired/target output current of a matrix converter operating in a switching state may be calculated for each switching state/state of the matrix converter.

Improved reliability may be obtained by determining at least three switch states based on a prediction error between the predicted output current and the desired output current.

In at least one embodiment, the selection of which switch states to use to generate the control strategy may be further narrowed based at least on the input current associated with each switch state and/or the output voltage associated with each switch state.

Identifying the plurality of switch states may include: obtaining a target input transformation result representing a mathematical transformation result of a desired input current of the matrix converter; for each possible switch state of the matrix converter, obtaining an input transformation result representing a mathematical transformation result of a current input of the matrix converter associated with the possible switch state; identifying a plurality of input transformation results that are adjacent to a location of the target input transformation result when mapped using a Cartesian coordinate system; and identifying a plurality of switch states associated with the identified plurality of input transformation results.

Embodiments thus recognize that the number of switch states required to calculate a predicted output current may be reduced based at least on a desired input current.

Each switch state may be associated with a respective current input. The mathematical transformation of the current input for each switch state may be used to identify certain switch states further based on, for example, a mathematical transformation of the desired current input.

Thus, embodiments enable generating a control strategy further based on a desired input current of the matrix converter.

Identifying the plurality of switch states may include: for each possible switching state of the matrix converter, obtaining a second output transformation result representing a mathematical transformation result of the voltage output of the matrix converter associated with the possible switching state; and identifying the plurality of switch states based on the magnitude of the second output transform result.

Embodiments enable reducing the number of switch states required to calculate a predicted output current based at least on the magnitude of the output voltage of the matrix converter. In particular, it is to be recognized that each switch state may be associated with a respective output voltage.

A mathematical transformation of the respective output voltage for each switch state may be used to further identify or narrow which switch states are to be used to generate the control strategy.

Thus, embodiments enable generating a strategy further based on the voltage output of the matrix converter, and in particular based on the magnitude of the voltage output by the matrix converter.

In at least one embodiment, identifying the plurality of switch states based on the magnitude of the second output transformation result includes identifying a plurality of switch states associated with the second output transformation result having the greatest magnitude.

Identifying the at least three switch states may include: identifying a first switch state in which the voltage difference between all output terminals of the matrix converter operating in accordance with the first switch state is substantially zero; identifying a second switch state in which a voltage difference between at least two output terminals of the matrix converter operating according to the second switch state is non-zero; and identifying a third switching state in which a voltage difference between at least two output terminals of the matrix converter operating according to the third switching state is non-zero.

Thus, embodiments may include identifying at least one zero switch state and one non-zero or active switch state.

Generating the control strategy may include: calculating duty cycles of the at least three switching states based on the target output transform result; and generating a control strategy for the matrix converter based on the calculated duty cycle.

Preferably, the mathematical transformation is an alpha-beta transformation, such that the target output transformation result represents an alpha-beta transformation result of a desired multiphase output current of the matrix converter, and each predicted output transformation result represents an alpha-beta transformation result of an output current predicted for a respective switching state.

A computer program is also presented, which is adapted to perform the method as described previously when run on a computer.

According to another embodiment of the invention, there is provided a modulation strategy generator for a three-phase to three-phase matrix converter, the matrix converter being operable in a plurality of switch states, each switch state being associated with a respective switch state, the modulation strategy generator comprising a processor adapted to: obtaining a target output transformation result representing a mathematical transformation result of a desired output current of the matrix converter; identifying a plurality of switch states of the matrix converter; for each of the identified plurality of switching states, obtaining a predicted output transformation result representing a mathematical transformation result of the output current predicted for the switching state; identifying at least three switch states from the plurality of switch states, wherein the location of the target output transformation is encompassed by an area defined by locations of predicted output transformations associated with the at least three switch states when mapped using a Cartesian coordinate system; and generating a modulation strategy for the matrix converter based on at least three switch states.

The modulation strategy generator may be adapted to obtain a predicted output transformation result from a simulation model of the matrix converter for each switching state.

The modulation strategy generator may be adapted to predict an output current of the matrix converter associated with each of the identified plurality of switching states using a simulation model of the matrix converter; and performing a mathematical transformation on the predicted output current associated with each switching state to obtain a predicted output transformation result for each of the identified plurality of switching states.

The modulation strategy generator may be adapted to: for each of the identified plurality of switch states, obtaining a predicted output transform result error representing a prediction error between a predicted output transform result associated with the switch state and the target output transform result; and identifying at least three switch states, wherein, when mapped using a cartesian coordinate system, the location of the origin is encompassed by an area defined by the locations of predicted output transform result errors associated with the at least three switch states.

Drawings

Examples of the invention will now be described in detail with reference to the accompanying drawings, in which:

fig. 1A, 1B, and 1C each show a matrix converter;

FIG. 2 is a flow diagram illustrating a method according to an embodiment;

FIGS. 3A and 3B illustrate mathematical transformations of output currents predicted for a plurality of switch states of a matrix converter;

FIG. 4 illustrates predicted output current vectors for a plurality of switch states of a matrix converter;

FIG. 5 illustrates a control strategy according to an embodiment;

FIG. 6A illustrates a mathematical transformation of a plurality of switch state current inputs for a matrix converter;

6B-6C each illustrate a mathematical transformation of a plurality of switch state voltage outputs for a matrix converter;

FIG. 7 illustrates a mathematical transformation of predicted output current errors for a plurality of switch states of a matrix converter;

FIG. 8 illustrates a control strategy generator according to an embodiment;

FIGS. 9A and 9B illustrate mathematical transformations of output current and current input, respectively, predicted for a plurality of switch states of a matrix converter;

FIG. 10 illustrates a system according to an embodiment;

fig. 11A and 11B illustrate experimental results of a system according to an embodiment.

Detailed Description

According to an embodiment of the invention, a method of generating a control strategy based on at least three switching states of a matrix converter is provided. The at least three switch states are selected based on at least the predicted output current and the desired output current associated with each switch state. In particular, a mathematical transformation of the output current and the output current associated with each of the plurality of switch states is desired for identifying the appropriate switch state.

Embodiments are based, at least in part, on an implementation that can generate a control strategy that achieves reliable responsive control of a matrix converter based on a predicted output current and a desired output current. Embodiments provide a reliable method of identifying suitable switching states of a matrix converter for generating a control strategy.

The illustrative embodiments may be used, for example, in an electrical drive or integrated drive using a matrix converter. Other implementation strategies for such a matrix converter will be apparent to the skilled person.

Fig. 1A, 1B and 1C each show a matrix converter 10 adapted to convert a three-phase input signal to a three-phase output signal according to various embodiments.

The matrix converter 5 comprises three input nodes 11 each connected to receive a respective phase of an input signal from a three-phase AC supply 12. The matrix converter further comprises three output nodes 13 each connected to provide a respective phase of the output signal to a load 14.

The voltage supply 12 may be, for example, a typical three-phase mains power supply or other three-phase AC power supply. The voltage supply 12 may be modeled, for example, as three voltage or current sources, each associated with a different phase. Each voltage or current source may be provided with an inductor and a damping resistor connected in parallel, each pair of inductor and damping resistor connecting the respective voltage or current source to the respective input node 11.

The load 14 may be, for example, a capacitive load or an inductive load, so that the matrix converter may comprise a capacitive port 15 as illustrated in fig. 1A or an inductive port 16 as illustrated in fig. 1B, or both.

It will of course be appreciated that in some embodiments, such as the one illustrated by fig. 1C, the matrix converter 1 need not include a particular output port. Such an embodiment may be used, for example, if the load 14 comprises an inductor.

Each output node 13 may be connected to each input node 11 through a respective bidirectional switch. The matrix converter 10 thus comprises nine (3 × 3) bidirectional switch arrays.

A capacitor arrangement 18 may be provided, as illustrated in fig. 1B, in order to provide a path for the inductor current of each phase. For example, if the matrix converter comprises a capacitive port 15, such as the capacitive port illustrated in fig. 1A, such a capacitor arrangement may not be required.

To prevent a line-to-line short (of the voltage source), the two bidirectional switches associated with a single output node should not be turned on at any given moment. Similarly, to ensure the path of the inductor current of each phase of the input signal via the capacitor arrangement 18 or the capacitive port 15, the output node 13 should not be disconnected from each input node 12. This prevents large overvoltages from occurring. In other words, each output node 13 must always be connected to receive the phase of the voltage source 12. Both of these limitations allow for improved safety, reliability and longevity of the device.

It will be apparent to those skilled in the art that the matrix converter 10 may operate in a limited number of switch states, each switch state representing a different on-off configuration of the bidirectional switch. For the matrix converter 10 of fig. 1, there are only 27 switch states that meet the above-mentioned limits.

As used herein, a "zero switch state" is defined as a switch state in which the voltage between each output node and a reference voltage is substantially the same. For example, each output node 13 may be connected to the same input node 11. Thus, there is no or negligible voltage difference between any of the output nodes 13 (since each output node is at the same voltage).

A "non-zero switching state" or an "active switching state" is defined as a switching state in which the voltage difference between each of the at least two output nodes and the reference voltage is different. For example, two or more output nodes may be connected to different input nodes. Thus, there is a voltage difference between at least two output nodes.

The control strategy may be used to define which output node is connected to which input node at any given time (i.e. in which switching state the matrix converter operates). This control strategy may enable modulation of the voltage supply to the load. In particular, a pulse width modulation control strategy may be used to define how long a matrix vector operates in a particular switching state.

A controller (not shown), such as a Field Programmable Gate Array (FPGA), may use the control strategy to control the switching of the bidirectional switches. By way of example only, the controller may provide a variable voltage connection to one or more transistors of each bidirectional switch to control the conductivity of the transistors to effect control of the bidirectional switches.

With further reference to fig. 2-4, a method 2 of generating a control strategy for a multiphase output matrix converter 10 will be described in accordance with an embodiment.

For the embodiments described below, the transform results generally refer to the results of an alpha-beta transform of the polyphase signal. Of course, other results of the mathematical transformation of the polyphase signal may be used to achieve various advantages, such as a direct orthogonal zero transformation, also known as dq0, dqo, 0dq or odq transformation. In general, the mathematical transform changes the reference frame of a particular polyphase signal and preferably provides a two-dimensional or two-part result in a manner similar to a complex value. The skilled person will appreciate that this result may be represented as a pair of numbers (e.g. coordinates), or by a corresponding vector.

Fig. 2 shows a flow chart of a method 2 for generating a control strategy according to an embodiment.

Fig. 3 and 4 each show the transformation results associated with the desired output current and the predicted output current of the matrix transformer, plotted using a cartesian coordinate system in a mathematical (e.g., a- β) plane.

The method includes obtaining 20 a target output transform result 30 representing a mathematical (e.g., alpha-beta) transform result of a desired multiphase output current of the matrix converter 10. When viewed as a vector from the origin, target output conversion result 30 may be viewed as a target output current vector.

The method also includes identifying 22 a plurality of switch states associated with the matrix converter 10. This may include, for example, identifying only a subset or selection of all available switch states of the matrix converter 10 for further processing. In other examples, all switch states associated with the matrix converter are identified for further processing.

Preferably, the identified plurality of switching states includes at least one zero switching state and two or more non-zero switching states. Even more preferably, the identified plurality of switching states includes at least one zero switching state and six or more non-zero switching states.

The method further comprises obtaining 24, for each identified switching state, a predicted

output transformation result

31, 32, 33, 34, 35, 36, 37 representing a mathematical (e.g. alpha-beta) transformation result of a predicted multiphase output current of the matrix converter 10 operating according to the respective switching state. When viewed as a vector from the origin, each predicted output transform result may be viewed as a predicted output current vector associated with the respective switch state.

Thus, fig. 3 and 4 may illustrate predicted output

current vectors

31, 32, 33, 34, 35, 36, 37 associated with predicted output currents for the identified plurality of switch states and target output current vector 30 associated with the target/desired output current.

The alpha-beta transformation result of the predicted output current (i.e. the predicted output current vector) may be obtained, for example, from a model or simulation of the matrix converter and an associated load, which may be a predicted load or a default load. For example, the model may include a data set or table, or in some embodiments, may include circuit simulation software. The output current may be considered to be the current provided to the load (i.e., the load current).

Equations (1) and (2) illustrate a predictive load model for a simple RL load (having resistance R and inductance L). The equation uses the switching period T _S (the inverse of the switching frequency).

e _j ＝I ^j _o (k+1)-I _o (k) (2)

Wherein, I _o (k + 1) and I _o (k) Is the load current for j = {0,1,2 \8230; } at (k + 1) and k instants, respectively, where j is the identified switching state.

It has been recognized that different switch states are associated with different predicted output currents, which are respectively represented by different output transformation results defining a predicted output current vector.

As shown in fig. 3A and 3B, the transform results 31, 32, 33, 34, 35, 36, 37 of the predicted output currents may form one or more skewed polygons, here single skewed hexagons with offset centers.

The method 2 further comprises identifying 26 at least three switch states, the predicted output transformation results 31, 32, 37 associated with the at least three switch states defining a region 40 or zone in which the target output transformation result 30 is located. Thus, the switch states associated with at least three predicted output

current vectors

31, 32, 37 defining the region 40 containing the target output transformation result 30 may be identified.

Accordingly, the method may include identifying 26 at least three switch states for generating a control strategy.

Identifying at least three switching states preferably includes identifying at least one zero switching state (i.e., associated with zero switching state transition result 37) and at least two non-zero switching states (i.e., associated with first non-zero switching state transition result 31 and second non-zero switching state transition result 32).

Although the zero switch state provides the same voltage between the output nodes of the matrix converter, the output current provided by the matrix converter operating according to the zero switch state may be non-zero (e.g., due to the nature of the load, such as inductance, resistance, impedance, post-electromagnetic field effect, etc.). This is best illustrated by fig. 3A, which identifies the mathematical transformation 37 of the predicted output current associated with a zero switch state as non-zero.

Specifically, as shown in fig. 3B, the region containing the target output conversion result may be defined in the following manner. The predicted output transform result 37 associated with a particular zero switch state (i.e., zero switch state transform result 37) may define the center of a circle of infinite diameter. The method may comprise identifying a first non-zero vector (associated with a first non-zero switch state transition result 31) and a second non-zero vector (associated with a second non-zero switch state transition result 32), wherein a sector of the circle, when bounded on one side by a line beginning at zero switch state transition result 37 and intersecting with first non-zero switch state transition result 31 and bounded on the other side by a line beginning at zero switch state transition result 37 and intersecting with second non-zero switch state transition result 32, defines an area in which target output transition result 30 is located.

As another example, the predicted output transform results 31, 32, 37 for the identified at least three switch states may define vertices, such as triangles, of the area that includes the target output transform result 30.

As yet another example, identifying at least three switch states may include identifying three predicted output transform results 31, 32, 37 that are closest to the target output transform result 30, and identifying associated switch states.

In this way, method 2 selects which switch states are to be used in generating the control strategy based on at least a mathematical transformation of predicted output currents for different switch state matrix converters and a mathematical transformation of desired output currents of the matrix converters.

The method 2 further comprises generating 28 a control strategy for a plurality of identified switch states. The generating 28 may include determining an appropriate duty cycle for each of the plurality of identified switching states and generating a control scheme based on the identified duty cycles.

Referring now also to fig. 4, and as briefly described above, each

transformation result

31, 32, 37 may be considered a predicted output current vector associated with a respective switch state or switch state.

By way of example, the first non-zero switch state transition result 31 may be related to a first predicted current vector

In association, a second non-zero switch state transition result 32 may be associated with a second predicted current vector ≦>

Is associated and the zero switch state transition result 37 can be associated with a zero predicted current vector>

And (4) associating. Similarly, target output transformation result 30 may correspond to a target current vector

And (4) associating.

Generating the control strategy may include identifying a linear combination of the first predicted current vector 51, the second predicted current vector 52, and a zero predicted current vector that yields the target current vector 50. The calculated duty cycle of each predicted current vector is assigned to the duty cycle of its associated switching state.

Specifically, an appropriate linear combination can be calculated by considering the following equation.

d ₁ +d ₂ +d ₀ ＝1 (4)

Wherein the content of the first and second substances,

represents a first prediction current vector, is greater than or equal to>

Represents a second prediction current vector +>

Represents a zero prediction current vector, is greater than or equal to>

Representing the target current vector, d ₁ Representing the duty cycle of the first predicted current vector, d ₂ Representing the duty cycle of the second predicted current vector, d ₀ Representing the duty cycle of the zero predicted current vector.

For clarity, vectors are synthesized

And &>

Shown in fig. 4. It is apparent that the combination of these three resultant vectors produces the target current vector 50.

In this way, the duty cycle (i.e., relative operating time) of each predicted current vector that produces the appropriate target vector may be calculated. A control strategy for a plurality of identified switch states may be calculated based on the determined duty cycles.

For example, referring to fig. 5, generating the control strategy may include generating a two-sided switching pattern of the matrix controller based on the calculated duty cycle of the predicted current vector.

And zero predicted current vector

Associated zero switch state v ₀ It can be calculated as the duty cycle d ₀ Into three parts, two of which are of equal size, one being twice as large as the other two. And the first predicted current error->

Associated first non-zero switching state v ₁ It can be calculated as the duty cycle d ₁ Divided into two parts on average. Analogously to the second prediction current error->

Associated second non-zero switch state v ₁ It can be calculated the duty cycle d ₂ Divided equally into two parts.

The divided portions of the duty cycle may be arranged in a pattern as shown in fig. 5.

The proposed embodiments enable a fast dynamic response without compromising the quality of the output signal or waveform of the matrix converter.

The proposed method can be used to predict the output voltage with improved accuracy, since the dynamics of the input filter means that the output voltage generated by other control strategies, such as SVM and the like, may not be accurate.

In some embodiments, the predicted current vector may be a predicted input current vector to enable control of the input current of the matrix converter.

More detailed embodiments of generating the control strategy will be described with reference to fig. 6A to 7.

In an embodiment, identifying the plurality of switch states of the matrix converter may include iteratively limiting or narrowing the selection of switch states to identify only a portion of all available switch states. Many possible methods may be used to narrow or limit this choice.

As mentioned previously, for the matrix converter 10 of fig. 1, there are only 27 switch states that meet certain limits. Of these switching states, six are considered to be rotating or synchronous switching states that provide only changes in magnitude and direction between the input node and the output node. For simplicity, in at least one embodiment, these switch states need not be considered. The remaining vector/switch states may be referred to as: 0 (i.e., zero switch state), ± 1, ± 2, ± 3, ± 4, ± 5, ± 6, ± 7, ± 8 and ± 9. It should be noted that there are three zero switch states.

The present invention recognizes that the number of switch states can be further reduced or otherwise reduced by limiting the switch states that meet certain criteria for input/output signals.

As an example, the number of switching states may be reduced based on a multiphase input current of the matrix converter and/or a multiphase output voltage of the matrix converter.

The matrix converter 10 operating according to a particular switching state may be associated with a respective input current vector representing the input current of the matrix converter operating in the particular switching state.

Similarly, a matrix converter 10 operating according to a particular switching state may be associated with a respective output voltage vector representing the output voltages of the matrix converter operating in the particular switching state.

That is, each switch state is associated with a respective input current vector and output voltage vector, as illustrated in fig. 6A and 6B.

Fig. 6A illustrates input current vectors for the matrix converter 10, each vector being a result of a mathematical (e.g., a- β) transformation of the input current of the matrix converter 10 associated with a respective switch state.

Fig. 6B illustrates output voltage vectors of the matrix converter 10, each vector being a mathematical (e.g., a- β) transformation result of the voltage output by the matrix converter 10 for each respective switching state.

The selection of the switch states may be limited to only those switch states for which the associated input current vector is adjacent to the desired input current vector. Thus, the desired input current associated with the desired input current vector may define which switch states are selected.

The selection of the switch state may be based on the magnitude of the associated output voltage vector.

For example, in some embodiments, only the switching states of the output voltage vector having the largest magnitude (e.g., the outermost output voltage vector of fig. 6B) may be selected. This ensures maximum power output of the matrix converter.

In other examples, only the switch state of the output voltage vector having the lowest magnitude may be selected. This may ensure increased control over the angle of the input current.

More than one method of limiting which switch states are identified may be used to obtain certain advantages.

In one case, referring to fig. 6A, the desired input current transformation result is located within the first sector 61. The switch states that can produce input current in the sector are 3, ± 6, ± 9, ± 1, ± 4 and ± 7. The plurality of switch states is first limited to these identified switch states, i.e. the initial set of 12 switch states.

FIG. 6C illustrates an output voltage vector for the above scenario, where the selection of switch states has initially been reduced to only select switch states 3, + -6, + -9, + -1, + -4 and + -7, each associated with an input current that can provide the desired input current. The distribution of the output voltage vector for each of these switch states is illustrated in fig. 6C.

In an embodiment, only those switch states associated with the largest magnitude output voltage vector (i.e., those voltage vectors located in the outer hexagon of fig. 6C) are selected.

Thus, by way of example, switch states ± 4, ± 7 and ± 1 may be selected as the identified plurality of switch vectors for further processing.

Embodiments recognize that the switch states of a matrix converter may be associated with a plurality of vectors representing different parameters of the matrix converter. As previously described, the switch states may be associated with a predicted output current vector (representing a predicted multiphase output current of the matrix converter operating according to the switch states), an input current vector (representing a multiphase current input of the matrix converter operating according to the switch states), and a voltage output switch vector (representing a multiphase voltage output by the matrix converter operating according to the switch states).

Embodiments also recognize that the identification of the control strategy generated for the switch states may be based on a plurality of vectors (characteristics) associated with each switch state.

According to a preferred embodiment, obtaining a predicted output transformation result (i.e. a predicted current vector) for each identified switching state comprises obtaining a predicted current error vector for each identified switching state.

The predicted current error vector may be represented by a mathematical (e.g., alpha-beta) transformation result for a prediction error between an output current of the associated switch state current matrix converter and a desired output current. Thus, the error between the predicted output current and the desired output current can be calculated for each relevant switch state. The mathematical transformation of this error may represent a predicted current error vector for the switch state.

In other or further embodiments, the predicted current error vector for a switch state may be modeled as a difference between a transformation result of the predicted output current associated with the switch state and a transformation result of the desired output current.

Thus, the predicted output transform result error may be considered to represent a prediction error between the predicted output transform result associated with the switch state and the target output transform result.

FIG. 7 illustrates a predicted current error vector e for a plurality of identified switch states ₁ 、e ₂ 、e ₃ 、e ₄ 、e ₅ 、e ₆ . Thus, fig. 7 illustrates the transformation results associated with the predicted multiphase output current error of the matrix converter for different switch states, which results are plotted in a mathematical (e.g., α - β) plane using a cartesian coordinate system.

According to an embodiment, the goal is to minimize or obtain a zero output current error. This target is thus the origin of the plane (0, 0). It will therefore be appreciated that the "target output transformation result" as described with reference to figures 2-5 is now embodied as the origin of a plane. That is, the target output transform result may be a point located at (0, 0).

In FIG. 7The explicit predicted current error vector comprises at least one zero switch state predicted current error vector e ₀ A predicted current error vector (for a zero switching state associated) and a plurality of non-zero switching state predicted current error vectors, each associated with a respective non-zero switching state. The further predicted current error vector comprises at least a first predicted current error vector e ₁ A second predicted current error vector e ₂ And a third e ₃ And fourth e ₄ Fifth e ₅ Sixth e ₆ A current error vector is predicted.

The control problem is to find a linear combination of at least three predicted current error vectors that will result in zero current error (i.e., the goal is to find the origin). This may be achieved by a linear combination of a zero switch state predicted current error vector and a non-zero switch state predicted current error vector.

Specifically, a solution exists if the target (origin) is located within a region formed by a zero switching state predicted current error vector and at least two non-zero switching state predicted current error vectors. If the target (origin) is located outside this region, it is considered an over-modulation condition and a different measure needs to be taken to solve its problem, described later.

In a particular embodiment, it may be determined that a solution exists if the target is within a region defined by a zero switch state predicted current error vector and two adjacent non-zero switch state predicted current error vectors. Predicting a current error vector e if a second non-zero switch state ₂ Is the predicted current error vector from the two nearest non-zero switch states to the first non-zero switch state ₁ Of the first non-zero switching state, the first non-zero switching state predicted current error vector e ₁ Can be considered adjacent to the second non-zero switch state predicted current error vector e ₂ 。

For each pair of adjacent non-zero switch state predicted current error vectors, a solution exists if the following conditions are satisfied:

(e _x -e ₀ )×(-e ₀ )·(e _y -e ₀ )×(-e ₀ )≤0 (5)

(e _x -e ₀ )·(-e ₀ )＞0 (6)

(e _y -e ₀ )·(-e ₀ )＞0 (7)

wherein e is _x Predicting one of the current error vectors for a non-zero vector, and e _y The current error vector is predicted for adjacent non-zero vectors.

Solution e of the embodiment shown in FIG. 7 _x ＝e ₁ And e _y ＝e ₂ 。

Selecting adjacent pairs of non-zero vector predicted current error vectors that satisfy these requirements (i.e., as e) _x And e _y ) Along with a zero switch state predicted current error vector e ₀ In order to generate a control strategy.

A linear combination of these vectors can then be obtained by solving the following system of linear equations to obtain the target (i.e., origin)

(e _xα -e _0α )·d ₁ +(e _yα -e _0α )·d ₂ ＝-e _0α (8)

(e _xβ -e _0β )·d ₁ +(e _yβ -e _0β )·d ₂ ＝-e _0β (9)

d ₁ +d ₂ +d ₀ ＝1 (10)

Wherein, d ₁ And d ₂ Predicting respective duty cycles of adjacent pairs of current error vectors for non-zero switching states satisfying the conditions of equations (5) - (7), and d ₀ Is the duty cycle of the zero-switching state predicted current error vector.

However, if d ₁ +d ₂ (> 1), this means that the target point (origin) is located at the current error vector e predicted by the non-zero switching state ₁ 、e ₂ 、e ₃ 、e ₄ 、e ₅ 、e ₆ (i.e., other predicted current error vectors) outside of the hexagon formed by the bounded region. In this case, the adjacent pairs e of current error vectors are predicted by predicting the current error vector at a non-zero vector _x 、e _y Inter modulation to achieve the attempt to get the target point (origin). That is, the method includes identifying a satisfaction, etcThese requirements of equations (5) - (7) yield a non-zero vector of the resultant vector closest to the origin predicting the duty cycle of the adjacent pair of current error vectors.

Depending on the error prediction, the underlying generation strategy yields accurate duty cycles and fixed switching frequency operation. Due at least to the fact that the control strategy is generated for the predicted output current, the control strategy will have a fast transient effect.

The proposed control method provides a fast dynamic response to changes in the desired or target output current of, for example, a matrix converter, with little impairment of the quality of the controlled waveform or output signal.

Also, steady state performance may be improved due to the modulation method included in the method. Thus, the combination of predictive control and appropriate modulation described herein results in good steady-state performance with fast dynamic response.

This embodiment is considered to be a method of direct predictive current-error vector control (DPCVC). The idea is proposed to consider the current error in its vector form in a conversion plane (e.g. the α β plane) as a cost function in order to calculate the duty cycle or the application time of the converter switching states. The goal is to minimize the load current error, making it equal to zero if possible. Therefore, the target point reached when the load current error is plotted is the origin of the plane.

Fig. 8 illustrates the control strategy generator 8 according to an embodiment. The control strategy generator 8 is adapted to generate a control strategy 87a for the matrix converter controller 88, which control strategy is adapted to control the matrix converter 89.

The control strategy generator 8 comprises a switch state supply unit 81 adapted to provide information on the switch state associated with the matrix converter 89. For example, the switch state supply unit 81 may indicate all available switch states of the matrix converter, or may only indicate non-rotating switch states of the matrix converter.

The control strategy generator 8 further comprises a switch state identification unit 82 adapted to identify a plurality of switch states of the matrix converter. For example, the switch state identification unit may identify a plurality of switch states based on a desired input current of the matrix converter or based on a magnitude of an output voltage vector associated with the switch states, as previously described with reference to fig. 6A-6C.

The control strategy generator 8 further comprises a circuit simulator 83 adapted to simulate the operation of the matrix converter 89. Specifically, the circuit simulator 83 includes models of an AC source 83a (which models the AC source of the matrix converter 89), an input filter 83b, a matrix converter 83c (which models the converter 89), and a load 83d (which models the load of the matrix converter 89). Predicted multiphase load current (i.e. output current) i _o Output by the circuit simulator.

Each of the plurality of switch states identified by the switch state identification unit is simulated by the circuit simulator. Thus, the model of the matrix converter 83c is controlled in accordance with each of the identified plurality of switch states.

Emulated multiphase output current i _o To the mathematical transformation unit 84 of the control strategy generator. The mathematical transformation unit 84 mathematically transforms (e.g., using an alpha-beta transform) the simulated multiphase current i _o A predicted output transformation result representing a mathematical transformation result of the output current predicted for the switching state is obtained for each of the identified plurality of switching states.

Furthermore, the mathematical transformation unit 84 may also obtain a target output transformation result representing a mathematical transformation result of the desired multiphase output currents of the matrix converter. This may be done, for example, by applying a desired input current i _o Performing a mathematical transformation, such as an alpha-beta transformation.

The switch state selection unit 85 identifies at least three switch states from among the plurality of switch states, wherein the location of the target output transformation result is encompassed by an area defined by the locations of the predicted output transformation results associated with the at least three switch states when cartesian coordinate system mapping is used. Accordingly, the switching state selection unit identifies at least three switching states based on the predicted output conversion result and the target output conversion result.

The duty cycle generator 86 determines the duty cycle for each of the at least three switching states, for example, using the previously described method.

The control strategy generator 87 generates a control strategy for the matrix converter 89 based on the determined duty cycles of the at least three switching states, for example using the previously described method. In an embodiment, the control strategy generator 87 arranges at least three switching states into a control mode based on the determined duty cycle.

In some embodiments, the mathematical transform unit 84 is based on the output current i predicted by the circuit simulator _o And the desired output current 84a to obtain a mathematical transformation of the predicted output current error.

Of course, it should be appreciated that in some embodiments, the circuit simulator 83 generates a predicted output current error associated with each switch state, for example, based on the desired output current 84a, such that the mathematical transformation unit 84 performs a mathematical transformation of the predicted output current error.

In an embodiment, the simulation or mathematical model (e.g., circuit simulator 83) may be a table or data set that identifies the predicted output current (or more preferably, the predicted output current error) for each switch state for a variety of different possible loads and/or input currents. Other simulation or mathematical models, such as circuit simulation software packages, will be apparent to the skilled person.

The method according to at least one embodiment may be iteratively repeated, since the predicted output current of the matrix converter may be dynamically changed according to e.g. changes to the load or current operation of the matrix converter. The load may react or respond in different ways to the matrix converter operating in different switching states, which may affect the prediction of the output current for the identified switching state.

Particularly advantageous embodiments may include continuously attempting to reduce the error in the output current by iteratively determining the output current error associated with the identified switch state and determining the control strategy based on the predicted output current error as previously described.

The predicted output current may be determined based on information about the present switch state and the present output current of the matrix converter. Thus, information about the current or ongoing operation of the matrix converter can be used to predict possible output currents for different possible switch state matrix converters. As an example, information about the current output current may provide information about the load characteristics of the matrix converter, which may be used to improve the simulation of the load.

Thus, the control strategy may be automatically adjusted and changed over a period of time.

According to at least another embodiment, a method of generating a control strategy capable of controlling input currents and output currents of a matrix converter is proposed.

In particular, a method according to an embodiment may include identifying at least five switch states for generating a control strategy.

In a manner similar to the previously described method, identifying at least five switching states may include iteratively selecting a particular switching state from all possible switching states based on a predicted output current, a desired input current, and/or a magnitude of an output voltage vector associated with the switching state.

Preferably, identifying at least five switch states comprises identifying at least three switch states according to the method described with reference to at least fig. 2-7. In particular, the identified at least three switching states may include a zero switching state and at least two non-zero switching states.

For purposes of the embodiments described below, the three identified switching states may include a zero switching state, a first non-zero switching state, and a second non-zero switching state.

As previously described, each switch state may be associated with a transformation of a predicted output current error, which corresponds to a predicted output current error vector. Thus, as illustrated in FIG. 9A, a zero switch state and zero predicted output current error vector e ₉₀ Correlating the first non-zero switch state with a first predicted output current error vector e ₉₁ Associated with a second non-zero switch state and a second predicted output current error vector e ₉₂ And (4) associating.

The method further includes identifying at least two other switch states, a third non-zero switch state and a fourth non-zero switch stateStatus. Third predicted output current error vector e associated with third non-zero switch state ₉₃ Error vector e of output current basically located at zero ₉₀ And a first predicted output current error vector e ₉₁ On the intersecting line. A fourth predicted output current error vector e associated with a fourth non-zero switching state ₉₄ Error vector e of output current basically located at zero ₉₀ And a second predicted output current error vector e ₉₂ On the intersecting line.

In this way, and as shown in FIG. 9B, with a first non-zero vector i _{i_1} And a second non-zero vector i _{i_2} The phase angles of the associated input current vectors are the same and are associated with a third non-zero vector i _{i_3} And a fourth non-zero vector i _{i_4} The phase angles of the associated input current vectors are the same. However, the magnitudes of the first and third non-zero input current vectors may be different, as may the magnitudes of the second and fourth non-zero input current vectors.

Once five switch states have been identified, including four non-zero switch states and one zero switch state, a control strategy for the matrix converter may be generated. The goal is to achieve control of the input current angle and maintain zero output current error. It is assumed that the desired input current is in phase with the voltage supplied to the matrix converter, i.e. from the voltage supply.

Referring specifically to fig. 9B, the selected four non-zero switch states are each associated with a respective input current vector representing the transformation result. Thus, the first non-zero switch state and the first input current vector i _{i_1} Associating a second non-zero switch state with a second input current vector i _{i_2} Associated with a third non-zero switching state and a third input current vector i _{i_3} Associated with a fourth non-zero switching state and a fourth input current vector i _{i_4} And (4) associating. The input current associated with the zero-switching state is 0, so that the input current vector for the zero-switching state is a zero vector at the origin.

Reference input current vector i _{i_ref} Can be taken from the angle b with the reference input current vector _i The voltage supply of (2) is obtained. For example, the reference input current vector may be obtained from a multi-phase voltage output by a voltage supply (since it may be assumed that the voltage output of the voltage supply should be synchronized with the input current of the matrix converter).

A linear combination of the five identified switch states that satisfies both the output current vector requirement and the input current vector requirement by solving the following system of linear equations may be identified:

(e _91β -e _90β )·d ₁ +(e _92β -e _90β )·d ₂ +(e _93β -e _90β )·d ₃ +(e _94β -e _90β )·d ₄ ＝-e _90β (12)

(-i _{i_1∝} sin(bi)+i _{i_1β} cos(bi))·d ₁ +(-i _{i_3∝} sin(bi)+i _{i_3β} cos(bi))·d ₃ ＝0 (13)

(-i _{i_2∝} sin(bi)+i _{i_2β} cos(bi))·d ₂ +(-i _{i_4∝} sin(bi)+i _{i_4β} cos(bi))·d ₄ ＝0 (14)

d ₁ +d ₂ +d ₃ +d ₄ +d ₀ ＝1 (15)

and d is ₁ ,d ₂ ,d ₃ ,d ₄ Is the duty cycle of each respective non-zero switching state, d ₀ Is the duty cycle of the zero switching state.

At least one embodiment provides a method of generating a control strategy using multiple switch states based on a single analysis of a desired output current. This control strategy provides a fast response to changes in the demanded output current with high accuracy and low total harmonic distortion. The control strategy may be iteratively generated to iteratively ensure that the appropriate current is output.

The control strategy described herein enables the provision of a fixed or predictable switching frequency of the matrix converter, which increases the ease of design of the input filter.

Although the embodiments described above relate to a three-phase to three-phase matrix converter, it should be understood that matrix converters according to other embodiments of the invention may be connected to input supplies of any number of phases, for example a two-phase supply or a four-phase supply. It is envisaged that embodiments may be applied to matrix converters adapted to output a transformed supply of any number of phases, for example two, four or five phase outputs.

The proposed embodiments enable the method to be modified (e.g. in saturation or over-modulation conditions) to prioritize the control of the input current and/or the output current. Specifically, the method may be modified to prioritize control of input current (i.e., rather than output current) as required. This may be performed by obtaining a predicted input current and its appropriate transformation result.

In some embodiments, a system may be provided comprising a processing arrangement adapted to perform any of the methods previously described with reference to fig. 1-7, 9A or 9B.

As an example, as shown in fig. 10, an embodiment may include a computer system 100. The components of computer system/server 101 may include, but are not limited to, one or more processing arrangements (e.g., including a processor or processing unit 101), a system memory 104, and a bus 108 that couples various system components including the system memory 104 to the processing unit 101.

Bus 108 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor bus or local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, micro Channel Architecture (MCA) bus, enhanced ISA (EISA) bus, video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

Computer system/server 100 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by computer system/server 100 and includes both volatile and nonvolatile media, removable and non-removable media.

The system memory 104 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 105a and/or cache memory 105b. The computer system/server 100 may also include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, the storage system 104 may be provided to read and write data from non-removable, nonvolatile magnetic media (not shown and commonly referred to as "hard drives"). Although not shown, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk such as a CD-ROM, DVD-ROM, or other optical media may be provided. In such cases, each may be connected to bus 90 by one or more data media interfaces. As will be further depicted and described below, memory 104 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.

A program/utility 107a having a set (at least one) of program modules 107b may be stored in memory 104, as an example and not by way of limitation, an operating system, one or more application programs, other program modules, and program data. Each operating system, one or more application programs, other program modules, and program data, or some combination thereof, may include an implementation of a network environment. Program modules 108b generally carry out the functions and/or methodologies of embodiments of the invention as described herein.

The computer system/server 100 may also communicate with one or more external devices 109a (e.g., keyboard, pointing device, display 109b, etc.), one or more devices that enable a user to interact with the computer system/server 100, and/or any device (e.g., network card, modem, etc.) that enables the computer system/server 100 to communicate with one or more other computing devices. Such communication may occur via an input/output (I/O) interface 102. Also, the computer system/server 100 may communicate with one or more networks (such as a Local Area Network (LAN), a general Wide Area Network (WAN), and/or a public network (e.g., the Internet) through the network adapter 103. As depicted, the network adapter 103 communicates with other components of the computer system/server 100 via the bus 108. It should be understood that although not shown, other hardware and/or software components may be used in conjunction with the computer system/server 100. Examples include, but are not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data archival storage systems, and the like.

Fig. 11A and 11B illustrate experimental results for a system having a three-phase output matrix converter controlled by a control strategy generated by a method as previously described with reference to at least fig. 1-6C and 8.

TABLE 1

Table 1 shows exemplary parameters of the system in which the experiment was performed. The filter inductance and capacitance are for the filter (e.g., input filter 83 b) that filters the AC source provided to the matrix converter (with supply voltage) previously described. The input filter consists of an LC filter, wherein a damping resistor (of a damping resistor) is connected in parallel with an inductor. An input filter may be required to attenuate switching frequency harmonics. The system has a load with a load inductance and a load resistance. The switching frequency indicates the switching frequency of the matrix converter.

Consider the scenario where the system requires a 5A, 30Hz load current, which results in a three phase output current, as illustrated in the first diagram 111 of fig. 11A. Specifically, voltage outputs associated with the first phase 111A, the second phase 111B, and the third phase 111C are illustrated. Fig. 11A also illustrates a second graph 112 showing the matrix converter output line voltage 112 and a third graph illustrating the harmonic spectrum 113 (one of the three phases).

The harmonic spectrum indicates harmonics at the switching frequency (12.5 kHz) and multiples thereof. This means that the fixed switching frequency is advantageously generated by the proposed control strategy. The Total Harmonic Distortion (THD) of the controlled waveform is about 3.97%.

To test the transient behavior of the control strategy, step requirements of amplitude (e.g. 2 to 5A) and frequency (e.g. 20-40 Hz) of the output current of the matrix converter may be required. That is, the desired output current of the matrix converter may be changed from a current of amplitude 2A, frequency 20Hz to a current of amplitude 5A, frequency 40 Hz.

The resulting load current waveform is shown in fig. 11B and indicates the fast transient response achieved by this method.

Specifically, fig. 11B indicates a step demand for the three-phase load current 114, which illustrates a fast response to a desired output current change. That is, the output current of the matrix device immediately responds to the change in the desired current.

Fig. 11B also shows a mathematical (e.g., α - β) conversion result 115 of the three-phase output currents of the matrix converter represented by the first line 115A and the second line 115B. Here, the first line 115A represents the instantaneous α portion of the transform result, and the second line 115B represents the instantaneous β portion of the transform result.

Due to the low total harmonic distortion, it can be appreciated that the described method provides a fast response control strategy for matrix converters with low harmonic distortion.

Embodiments described herein relate to a method of generating a control strategy for a multiphase output matrix converter operable in a plurality of switching states. Such methods include obtaining a desired multiphase output current of the matrix converter, obtaining a predicted multiphase output current of the matrix converter operating in a particular switching state, and determining a control strategy based on the desired multiphase output current and the predicted multiphase output current.

In a particular embodiment, a mathematical transformation of the desired output current and the predicted output current is used to identify at least three switch states for use in the control strategy.

Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. Other bi-directional switches than those explicitly disclosed herein will be known to those skilled in the art, for example, diode-bridged bi-directional switching cells. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims shall not be construed as limiting the scope.

Claims

1. A method of generating a control strategy for a multiphase output matrix converter, the matrix converter being operable in a plurality of switching states, the method comprising:

obtaining a target output transform result representing a mathematical transform result of a desired multiphase output current of the matrix converter;

identifying a plurality of switch states of the matrix converter;

for each of the identified plurality of switching states, obtaining a predicted output transform result representing a mathematical transform result of an output current predicted for the switching state;

identifying at least three switch states from the plurality of switch states, wherein, when mapped using a Cartesian coordinate system, the location of the target output transformation is encompassed by an area defined by locations of predicted output transformations associated with the at least three switch states; and

generating a control strategy for the matrix converter based on the at least three switch states,

wherein identifying the plurality of switch states comprises:

obtaining a target input transformation result representing a mathematical transformation result of a desired input current of the matrix converter;

for each possible switch state of the matrix converter, obtaining an input transformation result representing a mathematical transformation result of a current input of the matrix converter associated with the possible switch state;

identifying a plurality of input transformation results that are adjacent to a location of the target input transformation result when mapped using a Cartesian coordinate system; and

a plurality of switch states associated with the identified plurality of input transformation results is identified.

2. The method of claim 1, wherein obtaining a predicted output transform result comprises, for each of the identified plurality of switching states:

for each switch state, a predicted output transformation result is obtained from a simulated or mathematical model of the matrix converter and a load of the matrix converter.

3. The method of claim 1, wherein obtaining a predicted output transform result comprises, for each of the identified plurality of switching states:

predicting an output current of the matrix converter associated with each of the identified plurality of switch states using a simulated or mathematical model of the matrix converter and a load of the matrix converter; and

a mathematical transformation is performed on the predicted output current associated with each switching state to obtain a predicted output transformation result for each of the identified plurality of switching states.

4. The method of any of claims 1-3, wherein identifying the at least three switch states comprises:

for each of the identified plurality of switch states, obtaining a predicted output transform result error representing a prediction error between a predicted output transform result associated with the switch state and the target output transform result; and

at least three switch states are identified, wherein, when mapped using a Cartesian coordinate system, the location of the origin is encompassed by an area defined by the locations of predicted output transform result errors associated with the at least three switch states.

5. The method of any of claims 1-3, wherein identifying a plurality of switch states comprises:

for each possible switching state of the matrix converter, obtaining a second output transformation result representing a mathematical transformation result of the voltage output of the matrix converter associated with the possible switching state; and

identifying the at least three switch states based on the magnitude of the second output transformation result.

6. The method of claim 5, wherein identifying a plurality of switch states based on the magnitude of the second output transformation result comprises identifying a plurality of switch states associated with a second output transformation result of a maximum magnitude.

7. The method of any of claims 1-3, wherein identifying the at least three switch states comprises:

identifying a first switch state in which the voltage difference between all output terminals of the matrix converter operating in accordance with the first switch state is substantially zero;

identifying a second switching state in which a voltage difference between at least two output terminals of the matrix converter operating according to the second switching state is non-zero; and

a third switching state is identified in which a voltage difference between at least two output terminals of the matrix converter operating according to the third switching state is non-zero.

8. The method of any of claims 1-3, wherein generating the control strategy comprises:

calculating duty cycles for the at least three switching states based on the target output transition result; and

generating a control strategy for the matrix converter based on the calculated duty cycle.

9. A method according to any of claims 1-3, wherein the mathematical transform is an alpha-beta transform, such that the target output transform result represents an alpha-beta transform result for a desired multiphase output current of the matrix converter, and each predicted output transform result represents an alpha-beta transform result for an output current predicted for a respective switching state.

10. A computer-readable storage medium having stored thereon a computer program adapted to perform the method of any one of claims 1-9 when run on a processing device.

11. A control strategy generator for a multiphase output matrix converter, the matrix converter being operable in a plurality of switch states, each switch state being associated with a respective switch state, the control strategy generator comprising a processor adapted to:

identifying a plurality of switch states of the matrix converter by performing a process comprising:

identifying a plurality of switch states associated with the identified plurality of input transformation results;

for each of the identified plurality of switching states, obtaining a predicted output transformation result representing a mathematical transformation result of the output current predicted for the switching state;

generating a control strategy for the matrix converter based on the at least three switch states.

12. The control strategy generator of claim 11, wherein the processor is adapted to obtain a predicted output transformation result from a simulated or mathematical model of the matrix converter and a load of the matrix converter for each switch state.

13. The control strategy generator of any one of claims 11 and 12, wherein the processor is adapted to:

14. The control strategy generator according to any one of claims 11 and 12, wherein the processor is adapted to: